Crossover-controlled genetic algorithm and its application to inverse problem in SF _{6} interrupter

Abstract

From the application of genetic algorithm (GA) to the optimal design of some electromagnetic devices, it is found that its convergence speed is directly affected by the similarity of crossover codes. Based on the analysis for crossover operation a crossover-controlled genetic algorithm (CCGA) is presented. For increasing the interrupting capability at no load current, a permanent magnetic field is drawn into the arc-rotated interrupter of sulfur hexafluoride circuit breaker. Under the condition of same angular frequency at any point on arc, the desirable flux density is determined. On the basis of finite element analysis the optimal magnetic field distribution is obtained by applying CCGA. As a result 18.6 percent of CPU time taken by GA is saved and the interrupting performances at no load and short circuit currents are improved.